56 FRIDTJOF NANSEN. M.-N.Kl. 



At 50 metres the water should have moved during the same period 

 towards N55-'E (Nôy^E magnetic) with an average velocity of 3.6 cm. 

 per second. 



At 100 metres during the same period the water should have moved 

 towards NßS^E (N5o''E magnetic) with an average velocity of 3.4 cm. 

 per second. 



As the observations of the relative movements at 10 metres give such 

 very irregular values a progressive vector-diagram for this depth is not 

 constructed for the whole time of observation (see later). 



It seems astonishing that while the ice was moving with such a velo- 

 city towards S 66 "^ W, the water at 20 metres was moving northwards at 

 an angle of 108^ with the movement of the ice, and at 50 metres the 

 water was moving almost in the opposite direction of that of the ice (at 

 an angle of 169°), while at 100 metres the water was moving in a direc- 

 tion between that at 20 metres and that at 50 metres. 



If we are wrong in our supposition that the movements of the deepest 

 water (near the bottom) are so insignificant that they can be left out of 

 consideration, the above results would not be trustworthy. If, for instance, 

 the water of the deepest layers had been moving with a considerable velo- 

 city in the same direction as the ice-floe, the actual movement of the latter 

 would consequently be so much more rapid, and our progressive vector 

 diagram in Fig. 44 would have to be lengthened towards the right; the 

 direction and velocity of the movements at 20. 50, and 100 metres would 

 also have to be altered. But even then the movements at these depths 

 would be very much smaller than the movement of the ice-floe, and would 

 be directed, at great angles, towards the right. 



There seems, however, to be no reason why the movement of the 

 water at great depths — greater than 300 metres — should be considerable 

 in any south-westerl}' direction, considering that the water at 100 and at 

 50 metres, according to our observations, had such slow movements, and 

 those in north-easterly directions. The water at depths greater than 300 

 metres is comparatively warm water which must come from the south, and 

 evidently with very slow movement. We might thus expect that our re- 

 sults would give an approximate idea of the real movements of the water 

 at the different depths. 



A remarkable result of our observations is the great difference be- 

 tween the movement of the ice and the movement of the water at a depth 

 of 20 metres, considering that the ice had a great thickness, and conse- 

 quently its movement does not represent only the surface drift, but rather 

 the movement of the whole top layer of the sea, down to a certain depth 



